a Low-Power Autonomous Driving Testbed with ROS and TORCS and a pun based on Runs and Roses
- Ubuntu 18.04 Docker container (based on this image) with
- collect training data machine learning (image, sensor data, control data)
- implement and test own autonomous driving algorithms
- write bridges to existing autonomous driving frame works
- Low-end hardware requirements
- Batteries included (torcs / ros / bridges)
- Support of graphical desktop using any web-browser
- Update the GIT-submodules
git submodule init && git submodule update
- Run following script
run_environment.sh
. It builds the necessary docker image and starts the docker container. - open a browser and goto the url
http://127.0.0.1:6080/#/
- run
torcs
by using the desktop icon - Get to run the quickrace in the menu - you won't be able to though, it will lock expecting a car to connect from the ROS adapter
- Start the ROS torc bridge
roslaunch torcs_ros_bringup torcs_ros.launch
- Recent Linux Distribution (64bit)
- Docker CE
- 8 GByte RAM
- 6 Gigabyte of HDD/SSD
- OpenGL > 2.0 capable graphic card
- TORCS ROS and Patched TORCS - Florian Mirus
- Source code of this repository is licensed under MIT licence
- Base Docker Image - Apache license 2.0
- torcs-1.3.7 - GPL2 and other